A number of studies have been performed in an attempt to define display requirements for primary diagnosis in radiology. Most of these studies address resolution requirements for soft displays or different image- processing options which may improve observer performance. The results of these studies show that when primary diagnosis is performed on soft display, reader performance is lower than when performed on film for the detection of imaging features which contain high-frequency information. The issue of display luminance has only been mentioned and/or dealt with in a cursory way in these studies, however, and the combined effects of resolution and display luminance on observer performance has not been adequately examined. Based on our own data and those from similar efforts at other institutions, we propose to investigate the effect of display luminance on readers' performance for the detection of specific and often subtle abnormalities and to assess the combined effect of luminance and resolution in a comprehensive, multi-observer, multi- abnormality ROC study. The specific modes in our study design and the type of abnormalities we will study should enable us to partially define "brightness detail" requirements for the region of primary interest in electronic imaging as it relates to projection radiography in the clinical environment.